Supporting Information

Supporting Information

Supporting Information Chaboureau et al. 10.1073/pnas.1324002111 Validation of Paleoclimate and Paleobiome Simulations with the formation of evaporites. If now we look at the geo- Mesozoic Biomes Derived from LPJ. LPJ is a dynamic global vege- graphical distribution of the evaporites on our maps, one can tation model that computes photosynthesis, evapotranspiration, first conclude that they are less abundant than the coal deposits. and ultimately net primary production and surface cover of Second, evaporitic basins are mainly localized over the desert vegetation through 10 plant functional types (PFTs). These PFTs area, although some are found over the tropical biome. A closer are differentiated by physiological, morphological, phenological, inspection shows that most of the evaporites found over the bioclimatic, and fire response attributes (1). The phenology tropical biome are close to the tropical–desert transition area. differentiation is constructed referring to phenology of present- In detail, this transition is marked by savannah, which corre- day groups of plants: evergreen, raingreen, and summergreen. sponds to a highly seasonal climate (one rainy season and one Such a distinction is neither relevant nor very well constrained arid season). for Mesozoic vegetation, so here the woody PFTs have been As an intermediate conclusion, for each time interval, no big gathered according to their bioclimatic characteristics (tables mismatch appears between our bioclimatic maps and the data 1 and 2 in ref. 1), ultimately giving tropical, temperate, and record. This shows that (i) our numerical simulations can be boreal biomes. Temperate and tropical herbaceous PFTs were trusted for use as a basis to interpret the way the angiosperms kept in the analysis (Table S2). have colonized the temperate areas and (ii) 1,120 and 2,240 ppm scenarios produce better fit with continental data than 560 ppm. – Testing the Impact of pCO2: A Model–Data Comparison. The aim of To go a step further in our model data comparison and find this part is to test the 15 simulations performed with FOAM-LPJ out the best fit between 1,120 and 2,240 ppm for each time in- for the five continental configurations and for the three atmo- terval, we need to compare our climatic simulations with other spheric CO2 concentrations used here (560, 1,120, and 2,240 proxies, namely, reconstructed sea surface temperatures. Un- ppm). The five continental configurations stand for the Late fortunately, such datasets are not available for the Late Triassic Triassic, the Early Jurassic, the Early Cretaceous, the mid-Cre- and the Early Jurassic. Thus, for these periods we relied on taceous, and the Late Cretaceous. We first compare our model pCO2 estimates from the literature to choose our scenarios. outputs, e.g., biomes, to the geographical distribution of climate- These estimates have high uncertainties, and most rely on sto- sensitive sediments. Based on the works of Warren, Scotese, matal analyses of different tree species. The latest studies sug- Parrish et al., and Chumakov et al. (2–5), we have plotted the gest values of ca. 900 ppmv (7), whereas most studies agree on location of evaporites and coals on our five paleogeographic a background atmospheric pCO2 of ca. 1,000 ppm for the late maps (Fig. S1). Triassic and the early Jurassic [whereas a doubling of this value Coals are indicators of a humid climate and are found at low likely occurred at the Triassic–Jurassic boundary event (8–11)]. and high latitudes. They are not discriminating for the temper- Given this information and the good fit of coals with our simu- ature and can be found under cool or warm climatic conditions lation (Table S1), we chose 1,120 ppm as the best-fit scenario for (3). For the Late Triassic, very few sites characterized by the the late Triassic and the early Jurassic. presence of coal deposits exist. When computing the spatial fit Tentative reconstructions of sea surface temperature (SST) between coals and simulated biomes (i.e., the percentage of coal gradients are available for Cretaceous times, for which forami- 18 sites that are not found in an arid biome), it appears that at- nifera and fish tooth δ Op, TEX86, and D47 are used (12–20). mospheric CO2 concentrations of 1,120 and 2,240 ppm are Because of uncertainties concerning our understanding of the better to fit the data (Table S1). Indeed, some desert regions way marine organisms record temperature, an envelope de- appear at 560 ppm over the high latitudes where coal deposits limited by simulated winter and summer SSTs in the northern are found. For the Early Jurassic (180 Ma), many more sites with hemisphere is plotted for each simulation in Fig. S2. Temper- coal deposits have been listed. Here again, except for one site atures have been averaged between 60°W and 140°E because located at 35°S to 75°E, the agreement between model and data most SSTs estimates are coming from Tethysian locations. Be- is perfect for the high CO2 levels (1,120 and 2,240 ppm). An cause they are driving evaporation budget and related moisture atmospheric CO2 level of 560 ppm induces the extension of advection toward the continent, we pay particular attention to desert regions over the northern high latitudes where coal de- midlatitude to high-latitude SSTs. posits are found. For the Cretaceous geographies, the same con- For the Aptian, high-latitude SSTs have been derived from the clusions can be drawn. First, most of the coal deposits are clumped isotope thermometry D47 measured on Berrasian–late located outside the desert areas, and second, a CO2 concentra- Valanginian belemnites (12), which allows us to estimate tem- tion of 560 ppm does a less good job in simulating humid areas perature during shell biomineralization and to assess δ18O shells fitting the coal’s distribution. independent of seawater δ18O. Accounting for error bars, the Evaporites are generally used in the literature to constrain the authors suggest temperatures ranging from 10 to 20 °C at that areal extent of arid zones. However, based on the study of the time. At 1,120 ppm (Fig. S2A), simulated SSTs at the same South Atlantic evaporitic basin (visible on the Early Cretaceous latitude are clearly out of this range (0–7 °C), whereas the fit is map), Chaboureau et al. (6) have demonstrated that the evap- better at 2,240 ppmv, temperatures ranging from 4 °C to 17 °C orites deposited in the southern part of the Central segment (Fig. S2B). At low latitude, the congruence with the TEX86 data (20°S–10°S) may have been controlled by the climate favoring is also better with 2,240-ppm simulation, whereas temperatures aridity and high saline waters. In contrast, the evaporites of the are underestimated at 1,120 ppm. northern part (10°S to 5°N) can hardly be reconciled with the For 90 Ma, the temperatures have been reconstructed from climatic conditions occurring there and may be due to hydro- rudists (18), TEX86 measurements (14), and planktonic forami- thermal sources. This hypothesis is supported by the gradient nifera δ18O (19) and show strong variations. For instance, tem- found in the mineralogical composition from the north to the perature at 30°N range from 16 °C to 32 °C depending on the south. From this study, we also note that seasonally arid climate proxy used. The main uncertainties in the reconstruction of pa- is enough to simulate environmental conditions in agreement leotemperatures from biogenic carbonate include assumptions Chaboureau et al. www.pnas.org/cgi/content/short/1324002111 1of15 about the equilibrium fractionation in extinct species, about the Cretaceous (Fig. S2). Therefore, we have integrated and dis- true δ water value, and about the possible diagenesis. Here the cussed the results of these best-fit model–data simulations in our selected planktonic foraminifera data in the study of (19) ex- manuscript. clude altered carbon material, but the nonequilibrium calcifica- tion could cause underestimated actual upper ocean water Assessment of Model Limitations temperatures by the isotopic measurements. In contrast to the For consistency purposes, we used identical atmospheric CO2 foraminifera, the outer layer of certain groups of rudist is com- concentration (pCO2) for FOAM and LPJ in our analysis. pact and has a good potential to preserve the original chemical However, despite the common use of LPJ-like models (e.g., and isotopic composition (18). For both 1,120 and 2,240 ppm Biome4) to assess global vegetation changes under past and scenarios, our simulations fail to capture the lower range of low- future climate conditions, the behavior of such algorithms under latitude reconstructions (by the planktonic foraminifera). At pCO2 far higher than present, such as those of the Mesozoic, has 2,240 ppm, simulated temperatures are high enough at 60°N to not been studied to our knowledge. For example, the CO2 encompass reconstructed temperatures from foraminifera, which fertilization effect (CO2 rise that leads to an increase in range between 17 °C and 20 °C at this latitude (Fig. S2D) and are photosynthesis activity that ultimately drives a strong en- consistent with the high values at the low latitudes. The 1,120- hancement of net primary productivity) has been shown to be ppm experiment depicts temperatures that are too cool to match correctly simulated in temperate areas and overestimated in the data (Fig. S2C). tropical forests for present day (21), but to what extent this For 70 Ma, the robustness of the reconstructed temperatures effect and limitations are valid for Mesozoic vegetation and pCO2 18 from fish tooth δ Op (16) is more important because the oxygen remains unknown. To test this potential limitation and make sure isotopes of biogenic phosphate are less prone to postmortem that the vegetation changes discussed in the paper are linked to alteration (in comparison with biogenic carbonate) and no non- climate and not to an artificial fertilization effect, the climatic equilibrium oxygen isotope fractionation has been observed outputs from the FOAM experiments were used to force during precipitation of biogenic apatite.

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    15 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us